Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2017-04-20

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 timotimo masak: when i pass repeatedly, the AI doesn't win, it just keeps making two immediately neighbouring towers higher and higher :D
00:07 mcmillhj joined #perl6
00:08 timotimo Use of uninitialized value $github-user of type Any in string context.  /  in code  at /home/timo/perl6/install/share/perl6/site/sources/08A003E9585BF4F4FEBD6559AE70DDC410B29DDC (META6::bin)
00:08 timotimo i'm so good at this %)
00:25 mcmillhj joined #perl6
00:28 timotimo Sub object coerced to string (please use .gist or .perl to do that)
00:28 timotimo in any dump_children at /home/timo/perl6/install/share/nqp/lib/QASTNode.moarvm line 1
00:28 timotimo :\
00:32 timotimo No such source rod 'left to right left to middle right to middle left to right middle to left middle to right left to right left to middle right to middle right to left middle to left right to middle left to right left to middle right to middle'
00:32 timotimo :D
00:36 geekosaur hanoying?
00:36 skids joined #perl6
00:37 timotimo yup
00:37 timotimo List $source = $(((("left", "to", "right"), ("left", "to", "middle"), ("right", "to", "middle")), ("left", "to", "right"), (("middle", "to", "left"), ("middle", "to", "right"), ("left", "to", "right"))), ("left", "to", "middle"), ((("right", "to", "middle"), ("right", "to", "left"), ("middle", "to", "left")), ("right", "to", "middle"), (("left", "to", "right"), ("left", "to", "middle"), ("right", "to",
00:37 timotimo "middle"))))
00:37 timotimo List $target = $(((("middle", "to", "left"), ("middle", "to", "right"), ("left", "to", "right")), ("middle", "to", "left"), (("right", "to", "middle"), ("right", "to", "left"), ("middle", "to", "left"))), ("middle", "to", "right"), ((("left", "to", "right"), ("left", "to", "middle"), ("right", "to", "middle")), ("left", "to", "right"), (("middle", "to", "left"), ("middle", "to", "right"), ("left", "to",
00:37 timotimo "right"))))
00:37 timotimo that is not a rod that is on the stack!
00:40 timotimo i should go to bed
00:40 timotimo i can continue this tomorrow
00:46 perlawhirl joined #perl6
00:56 mcmillhj joined #perl6
00:57 samcv what am i seeing here
01:01 Geth ¦ specs: 63686a0042 | (Samantha McVey)++ | S22-package-format.pod
01:01 Geth ¦ specs: Update META info for the license section
01:01 Geth ¦ specs:
01:01 Geth ¦ specs: Use the SPDX license identifier under license, which is the most widely
01:01 Geth ¦ specs: used license tags meant for interoperability.
01:01 Geth ¦ specs:
01:01 Geth ¦ specs: If the license does not have an identifier, then it is recommended to
01:01 Geth ¦ specs: add a URL to the license under the support key.
01:01 Geth ¦ specs: review: https://github.com/perl6/specs/commit/63686a0042
01:01 Geth ¦ specs: 07a57f3cae | (Samantha McVey)++ (committed using GitHub Web editor) | S22-package-format.pod
01:01 Geth ¦ specs: Merge pull request #120 from samcv/package
01:01 Geth ¦ specs:
01:01 Geth ¦ specs: Update META info for the license section
01:01 Geth ¦ specs: review: https://github.com/perl6/specs/commit/07a57f3cae
01:04 Cabanossi joined #perl6
01:06 gdonald joined #perl6
01:32 awwaiid joined #perl6
01:37 Geth ¦ specs: dce35d18d5 | (Samantha McVey)++ | S22-package-format.pod
01:37 Geth ¦ specs: META6.json describe how to denote no license or no assertion
01:37 Geth ¦ specs: review: https://github.com/perl6/specs/commit/dce35d18d5
01:38 samcv very good. covered all the likely cases now :)
01:49 Cabanossi joined #perl6
02:08 noganex_ joined #perl6
02:09 samcv hehe
02:09 samcv bag(emulates(10), test-depends(301), version(813), name(813), api(4), description(813), excludes(9), resources(145), build-depends(195), supersedes(9), production(10), support(211), source-uri, licence(4), author(285), source-url(779), superseded-by(9), resource(13), tags(149), license(108), auth(155), perl(739), provides(803), meta6(54), history, repo-type(31), authors(411), owner(9), authority(8), depends(771), source-type(69))
02:10 samcv so we have 4 modules with misspelled license (they spelled it as licence)
02:10 samcv and 108 of our modules have licenses plus uh. whatever those 4 that misspelled them..
02:16 TimToady well, that's the british spelling, they spell 'defence' differently as well
02:16 TimToady and a couple other three things...
02:22 Geth ¦ ecosystem: e6a2236f61 | (Zoffix Znet)++ (committed using GitHub Web editor) | META.list
02:22 Geth ¦ ecosystem: s/META.info/META6.json/
02:22 Geth ¦ ecosystem: review: https://github.com/perl6/ecosystem/commit/e6a2236f61
02:24 baest joined #perl6
02:25 samcv ok my json is fixed
02:51 baest joined #perl6
03:01 samcv yay i can't use the meta6 bin because
03:01 samcv ===SORRY!===
03:01 samcv Failed to parse „~/.gitconfig“.
03:01 samcv :(
03:01 samcv won't even start.... ugh
03:04 Cabanossi joined #perl6
03:04 rindolf joined #perl6
03:32 rindolf joined #perl6
03:47 Cabanossi joined #perl6
03:58 baest joined #perl6
04:05 u-ou How do you do a runtime-mixin of an already created object? Is it just $s = $s but role?
04:06 TimToady that will clone $s
04:06 g0d355__ joined #perl6
04:06 TimToady use '$s does role' for modification in place
04:06 u-ou ahhh
04:06 u-ou thanks
04:16 wamba joined #perl6
04:17 TimToady joined #perl6
04:21 mcmillhj joined #perl6
04:45 Actualeyes joined #perl6
04:57 kurahaupo joined #perl6
05:02 Cabanossi joined #perl6
05:03 mcmillhj joined #perl6
05:26 gdonald joined #perl6
05:30 mcmillhj joined #perl6
05:32 jroth joined #perl6
05:37 CIAvash joined #perl6
05:38 xtreak joined #perl6
05:40 lowbro joined #perl6
05:40 lowbro joined #perl6
05:46 nadim joined #perl6
05:47 Cabanossi joined #perl6
05:51 RabidGravy joined #perl6
06:02 mcmillhj joined #perl6
06:09 Woodi hallo #perl6 :)
06:11 Woodi some days ago I read that Prolog requires first class backtracking. we have it in v6 ? also I learned logic programming is superset of functional programming... but not saying I understand all of this :)
06:13 gdonald joined #perl6
06:15 TimToady backtracking is pretty easy; it's unification that's hard
06:18 Cabanossi joined #perl6
06:21 TimToady the hardest part of logical programming, though, is managing your lexicon such that words always mean the same thing everywhere, which they usually don't in real life :)
06:23 TimToady funny how real life actually means "complex"...
06:23 TimToady but the opposite of "imaginary"...
06:24 wamba joined #perl6
06:32 darutoko joined #perl6
06:43 domidumont joined #perl6
06:47 mr-foobar joined #perl6
06:48 domidumont joined #perl6
06:48 El_Che TimToady: hehe
06:48 yoleaux 18 Apr 2017 19:52Z <Zoffix> El_Che: you're good for packages. 2017.04.2 is out
06:48 El_Che will start
06:51 domidumont joined #perl6
06:52 mst TimToady: without continuations it's non-trivial to push backtracking down into the runtime though
06:52 mst of course the WAM doesn't really do it that way anyway
06:56 mst Woodi: first class backtracking at the VM level is pretty rare, and not at all a requirement to implement prolog
06:56 mst Woodi: hence why there's more than one implementation of prolog in perl5
06:56 mst Woodi: also, please call perl6 perl6, not 'v6'
06:57 sunnavy joined #perl6
06:58 u-ou yay my perl6 thing works
06:59 samcv whew. just finished PR to META6 module and the META6::Test modules
07:00 samcv once accepted META6::Test will be able to check license names against the full list of standardized names, and if it's not correct, tell the user to fix it, or if license isn't on the list, if you set a
07:00 samcv URL for the license text then it will also be able to pass
07:01 samcv and if you put NONE or NOASSERTION it will give a little warning to remind you that's nonideal
07:02 samcv https://github.com/jonathanstowe/Test-META/pull/21 can we call it the Licensining? or uh. idk something catchier
07:04 llfourn joined #perl6
07:05 labster joined #perl6
07:06 gfldex samcv: it's JSON, no license should be null, not "NONE"
07:06 samcv what?
07:06 samcv null does not convey any actual afirmative information thuogh
07:07 samcv whether there is none or if it's just an empty field
07:07 samcv though we could depart from SPDX on that i guess? hm
07:07 samcv or take NONE or null as both being no license?
07:08 gfldex null indicates the absence of a value. "NONE" may very well be an acronym.
07:09 samcv abcense of a value is different from no license in the affirmative though
07:09 samcv just how NOASSERTION means that there could be a license but the creator of the meta file does not know and makes no assertion
07:09 samcv because no license and no assertion are distinct things
07:10 samcv if anything License being null would be closer to NOASSERTION
07:10 nadim joined #perl6
07:13 gfldex samcv: I disagree, NOASSERTION would be indicated by not having the field in the JSON file
07:13 gfldex not sure if META6.pm6 is handleing that properly tho
07:14 samcv i we could define NOASSERTION is no license key and that license: null means NONE and they are equivilant. that would be okay with me i guess.
07:14 samcv as long as there is a way to distinguish between those two. that is the important part. and that they are defined
07:15 samcv i still don't like license: null.... but regardless. it still is fine to have it mean SOMETHING which is better than meaning nothing i think
07:15 gfldex aye
07:15 samcv thanks for bringing that up gfldex
07:16 gfldex besides, if there is no license it's rather pointless to put it into the ecosystem to start with
07:16 samcv well. most modules have no license....
07:16 gfldex as nobody knows if the module can be used at all
07:16 parv joined #perl6
07:16 samcv so
07:16 samcv hopefully we can make issues with all of those modules to add a license. but it's a big undertaking
07:17 samcv when only about 10% of them have one in the META
07:17 gfldex they don't got a license because it was never enforced
07:17 samcv well yes
07:17 gfldex I may be able to automate that problem away
07:17 samcv with travis?
07:17 gfldex githubs api is rather nice
07:17 samcv oh? to notify?
07:17 samcv that would be nice
07:18 gfldex but first, $dayjob
07:18 samcv though it's useful now that i have META6::Test being able to check validity of license type things. may eventually want to just cause an error when no license is there. but it doesn't have to happen overnigtht. we're working toward licenses and that's the important part
07:18 samcv nice talking to you gfldex
07:18 jian joined #perl6
07:19 gfldex yw
07:19 jian HELP
07:19 gfldex WHAT?
07:19 gfldex no wait, I can answer that myself
07:20 gfldex m: "HELP".WHAT.say
07:20 camelia rakudo-moar 894ba8: OUTPUT: «(Str)␤»
07:20 samcv hi jian
07:20 jian hi.
07:20 jian hmm..
07:20 samcv please let me try and help
07:20 samcv ZofBot, HELP
07:21 jian I don't actually know what it is
07:21 samcv oh :(
07:22 gfldex can you provide some details of your undescribable problem please?
07:23 jian that is human or what ?
07:24 gfldex humans are an undescribable problem indeed :)
07:24 samcv humans are homo sapiens and are sentient beings who populate this channel in addition to the IRC bots here
07:24 samcv see https://en.wikipedia.org/wiki/Human
07:24 moritz humans are meat bags. There, I described the problem for you.
07:24 samcv https://en.wikipedia.org/wiki/IRC_bot
07:25 samcv AFAIK those are the only two things in this channel
07:25 gfldex the IRC bots are smarter tho
07:25 u-ou I'm not sure what the difference is
07:25 u-ou but I think I'm a bot
07:25 u-ou just sayin'
07:26 mst moritz: this is where I turn up with an emu, right?
07:26 Actualeyes joined #perl6
07:26 jeek joined #perl6
07:28 moritz mst: an emu_lator? :-)
07:28 mst I was going for a flightless biped joke, but you do you
07:29 jian maybe I should go =-=
07:29 samcv no!
07:29 samcv don't go
07:30 zakharyas joined #perl6
07:30 jian but I don't know how
07:31 jian close the tab is ok?
07:31 samcv no
07:31 samcv i want to help!
07:32 jian help what
07:33 samcv <jian> HELP
07:33 samcv help you
07:33 samcv you need help
07:33 gfldex samcv: don't be so selflessly selfish!
07:33 samcv i will help
07:33 samcv XD
07:33 jian ok
07:33 jian just tell me
07:34 jian waht is it
07:34 samcv ask your question again please
07:34 jian or what's this
07:34 moritz an IRC channel!
07:35 jian IRC?
07:35 moritz that was easy. Next customer!
07:35 samcv this in an IRC channel. this is the #perl6 room where we talk about the perl6 programming languages
07:35 moritz jian: https://en.wikipedia.org/wiki/Internet_Relay_Chat
07:36 gfldex I wonder if I should use subparse in Git::Config.
07:37 jian maybe I know something :-)
07:37 gfldex samcv: PR welcome (read: I'm not good with grammars.) https://github.com/gfldex/perl6-git-config/blob/master/lib/Git/Config.pm6
07:38 dakkar joined #perl6
07:39 abraxxa joined #perl6
07:40 samcv what do i do with this
07:40 samcv oh
07:41 samcv yeah it doesn't even work if i delete my .gitconfig
07:41 samcv so idk how to make this thing work
07:42 samcv not sure why META bin needs it tho :\
07:42 samcv why does it care about my git config
07:42 gfldex samcv: try meta6 --create-cfg-dir
07:43 gfldex samcv: it needs your github username and your e-mail to generate README.md
07:43 gfldex and a few other things
07:44 parv gfldex: why is there a specific need for anything GitHub when the module name is more generic?
07:45 gfldex samcv: at some point it will ask nicely for your github PW and do anything for you. :)
07:45 gfldex parv: you can skip the github part but if you don't any new module will be pushed automatically.
07:46 parv "github" part should be easily switchable
07:46 gfldex i wanted to add automatic travis support too but failed due to travis wonky api
07:46 gfldex and so we read: "meta6 --new-module --name=<Module::Name::Here> --force --skip-git --skip-github"
07:47 gfldex meta6.cfg needs to learn a new trick it seams
07:47 * gfldex takes notes
07:48 ufobat joined #perl6
07:50 samcv gfldex, anything? :-D
07:51 gfldex the goal is to telepathically sense the general intend of the programmer and do anything automatically indeed
07:51 gfldex the same is true for Perl 6 I believe
07:55 wamba joined #perl6
07:56 * gfldex self.dayjob()
07:56 xtreak joined #perl6
08:00 xtreak joined #perl6
08:01 g4 joined #perl6
08:13 eyck joined #perl6
08:15 xtreak joined #perl6
08:15 xtreak joined #perl6
08:17 samcv u: end
08:17 unicodable6 samcv, U+02FA MODIFIER LETTER END HIGH TONE [Sk] (˺)
08:17 unicodable6 samcv, U+02FC MODIFIER LETTER END LOW TONE [Sk] (˼)
08:17 unicodable6 samcv, 363 characters in total: https://gist.github.com/a2197e703ec67db1f7002f1251c2aabe
08:19 eyck joined #perl6
08:29 rindolf joined #perl6
08:31 jonas1 joined #perl6
08:36 u-ou roles!
08:36 eyck joined #perl6
08:43 xtreak joined #perl6
08:50 u-ou hi
08:52 Actualeyes joined #perl6
08:58 astj joined #perl6
09:03 Cabanossi joined #perl6
09:12 ChoHag joined #perl6
09:25 mcmillhj joined #perl6
09:41 mcmillhj joined #perl6
09:42 bjz joined #perl6
09:45 eiro joined #perl6
09:48 Cabanossi joined #perl6
09:48 cpage_ joined #perl6
09:50 xtreak joined #perl6
09:57 mcmillhj joined #perl6
10:13 astj joined #perl6
10:13 astj joined #perl6
10:17 notbenh joined #perl6
10:17 mcmillhj joined #perl6
10:18 Cabanossi joined #perl6
10:30 xtreak joined #perl6
10:32 mcmillhj joined #perl6
10:33 bjz_ joined #perl6
10:41 astj joined #perl6
10:46 gdonald joined #perl6
10:50 astj joined #perl6
10:52 Geth ¦ ecosystem: 72de400aaa | (Zoffix Znet)++ (committed using GitHub Web editor) | META.list
10:52 Geth ¦ ecosystem: .info -> .json
10:52 Geth ¦ ecosystem: review: https://github.com/perl6/ecosystem/commit/72de400aaa
10:53 mcmillhj joined #perl6
10:56 xtreak joined #perl6
11:03 Cabanossi joined #perl6
11:08 kurahaupo joined #perl6
11:27 noganex joined #perl6
11:29 mcmillhj joined #perl6
11:30 tadzik joined #perl6
12:00 TimToady joined #perl6
12:01 timotimo masak: fixing Adventure::Engine and Crypt is a fascinating experience in "how have lists changed in the last years"
12:01 timotimo also, if you %foo.push(%foo.invert), be ready to see some duplicates, because invert is lazily being iterated and is updating as the hash gets stuff pushed into it, lol.
12:08 Geth ¦ ecosystem: scmorrison++ created pull request #322: AWS::Pricing META.info > META6.json
12:08 Geth ¦ ecosystem: review: https://github.com/perl6/ecosystem/pull/322
12:09 xtreak joined #perl6
12:12 zakharyas joined #perl6
12:14 Geth ¦ ecosystem: c143a14304 | (Sam Morrison)++ (committed by Zoffix Znet) | META.list
12:14 Geth ¦ ecosystem: AWS::Pricing META.info > META6.json (#322)
12:14 Geth ¦ ecosystem: review: https://github.com/perl6/ecosystem/commit/c143a14304
12:17 domidumont joined #perl6
12:17 Cabanossi joined #perl6
12:23 japh__ joined #perl6
12:32 mcmillhj joined #perl6
12:32 MasterDuke_ joined #perl6
12:35 japh__ I'm getting "Cannot find method 'setlang' on object of type Perl6::HookGrammar" when I run perl6-debug-m (Rakudo 2017.04.2, MoarVM version 2017.04). I saw some references to this #perl6-dev on 2/28 and #perl6 on 3/21, but nothing that's a clear solution
12:39 japh__ I was getting the same set of errors under Rakudo 2017.03, so upgraded MoarVM, NQP and Rakudo to the latest tagged releases, but still no luck
12:39 jnthn I think timotimo++ was taking a look at it quite recently
12:44 nadim hi all, do we have something like Devel::Size?
12:50 japh____ joined #perl6
13:00 mcmillhj joined #perl6
13:04 timotimo japh__ there is a trivial fix for that error but then you get a debugger that fails at everything anyway
13:04 wamba joined #perl6
13:04 timotimo nadim, would that give you thelayout of your classes in memor or something?
13:05 timotimo i wanted something like pahole forp6opaque in the past
13:05 timotimo oh gotta go
13:07 jnthn I think the closest we have is the heap analyzer
13:23 cdg joined #perl6
13:29 salva joined #perl6
13:37 skids joined #perl6
13:40 beginner joined #perl6
13:53 gdonald joined #perl6
13:58 mcmillhj joined #perl6
13:59 beginner which is the best IDE for perl6
14:01 eroux joined #perl6
14:03 El_Che beginner: atom, vim, emac (non of which are IDE's, I know)
14:06 AlexDaniel joined #perl6
14:13 moritz isn't it ironic that, being immaterial, atom doesn't actually contain atoms?
14:13 moritz while most things that contain atoms don't have atom in the name
14:18 Cabanossi joined #perl6
14:30 committable6 joined #perl6
14:34 khw joined #perl6
14:46 timotimo moritz: atomoritz?
14:51 SmokeMachine m: my @a = :1a, :2b, :3a, :4a; say [⊎] @a
14:51 camelia rakudo-moar 277b6e: OUTPUT: «bag(b(2), a(8))␤»
14:52 SmokeMachine m: my @a = :1a, :2b, :3a, :4a; say Bag.new ⊎ @a
14:52 camelia rakudo-moar 277b6e: OUTPUT: «bag(b(2), a(8))␤»
14:59 nadim timotimo: Data::Dump::Tree can give the layout, it can show the "perl address" but it  can't show how much memory each element uses. In the P5 version I used Devel::Size to get that information. I wonder if there is some equivalent functionality in P6.
15:00 timotimo it can not give the kind of layout i'm talking about
15:00 El_Che aaaagh. Now that I build 9 different rakudo packages I hit travis' 50 minute limit
15:06 iyra joined #perl6
15:07 iyra does anyone have knowledge of the SCGI module? I'm trying to use the example in the readme here (first under 'Usage') https://github.com/supernovus/SCGI
15:08 iyra however I'm getting "Nothing given for new socket to connect or bind to" when I try to use it, seems to be caused by connect method, here: https://github.com/supernovus/SCGI/blob/master/lib/SCGI.pm6#L22
15:08 skids m: say Bag.new ⊎ :e_with_oatmea
15:08 camelia rakudo-moar 277b6e: OUTPUT: «bag(e_with_oatmea)␤»
15:13 timotimo iyra: how does your code look? did you SCGI.new with a :port?
15:13 iyra yes, I used   my $scgi = SCGI.new( :port(8118) );
15:14 kurahaupo joined #perl6
15:14 iyra however I tried with  my $scgi = SCGI.new( :host<localhost>, :port(8118) ); and I get the same error
15:17 Cabanossi joined #perl6
15:17 timotimo interesting
15:18 timotimo ugh, scgi doesn't have any tests
15:20 iyra using IO::Socket::INET.new in general seems to work, though (using the first example in the docs for it) https://docs.perl6.org/type/IO::Socket::INET
15:21 iyra i can't think why it's not working with scgi though
15:22 timotimo it could just have bitrotted
15:23 iyra hmmm
15:24 timotimo just running one of the examples gives me errors
15:25 eroux joined #perl6
15:25 timotimo oh
15:26 iyra here's the code and full error that i got: http://fh.tl/CO
15:26 timotimo i know what's wrong
15:27 timotimo it's passing :listen(1), which upsets it
15:27 timotimo it must be just :listen or :listen(True)
15:28 iyra what would be the best way to make that modification locally on my computer?
15:28 iyra I'm not familiar with perl modules :)
15:28 timotimo hold on
15:29 timotimo i'd like to know why the code had \ in these weird places to begin with
15:30 timotimo ah, it uses psgi, let's see.
15:31 timotimo ok, i think my fix is ifne
15:31 timotimo fine*
15:33 timotimo you have zef, right?
15:33 iyra yes
15:34 timotimo zef can install from git links, so try this:
15:34 timotimo zef install https://github.com/timo/SCGI.git
15:34 timotimo might need --force because i didn't change the version, actually
15:34 iyra do I have to remove the original scgi?
15:35 titsuki hi #perl6, I couldn't send a bug report to the RT system with email. Is there something trouble with the RT today ?
15:35 timotimo now i've bumped thev ersion
15:36 timotimo titsuki: it's supposed to work. only the ticket submission thing on the rt website has been disabled long ago because of spam
15:36 timotimo iyra: you do not, it should replace it
15:36 timotimo also, if you do it now, i actually did bump the version
15:37 titsuki timotimo: thx. I'll try it again.
15:41 iyra timotimo, that's strange, I'm still getting the same error, even after changing :listen(1) to just :listen
15:42 timotimo did zef install --force give any errors or something?
15:47 iyra no, it installed fine
15:47 iyra I cloned the repo and made the change, then tried again with --force and zef didn't give any errors
15:49 timotimo huh
15:49 timotimo can you put some very obvious thing in there, like a "say 'hello'" near the problematic code?
15:49 iyra i'll give it a go
15:51 timotimo if you want to make sure that code definitely gets used, just -I ../path/to/SCGI/lib/
15:52 aindilis joined #perl6
15:54 iyra so I've found that even after doing zef install with my new git repo, it still uses the old sources
15:54 iyra and using -I ../path/to/SCGI/lib/ doesn't seem to get it done either, it still references the old sources in my .rakudobrew directory
15:54 iyra i checked by looking at the source file it was using
15:58 timotimo oh interesting
15:58 timotimo it could be it just grabs the META.info which has source-url pointing at the original repo
15:58 iyra oh, good point, i'll change that
15:59 iyra in my fork, i changed the version number to 2.4, but zef install is still reporting that it's installing 2.3 https://github.com/iyra/SCGI/blob/master/META6.json
16:01 timotimo are you doing "zef install --force ."?
16:02 iyra no, I was using the my github URL; doing the command you just posted installed my new version, though
16:03 iyra oh, nice! it works
16:03 timotimo ah
16:03 iyra the :listen thing was the problem, not sure why it wasn't getting the new version from github though
16:03 iyra thanks a lot for your help
16:03 timotimo you're quite welcome
16:03 timotimo i opened a pullrequest on SCGI, too
16:04 iyra thanks, would be nice to see it fixed
16:04 timotimo turns out there was an issue posted about this back in march
16:04 timotimo but it didn't look like anybody looked for the solution
16:04 timotimo i just now also fixed two other modules :)
16:04 timotimo Adventure::Engine and Game::Crypt
16:04 TimToady is there an easy way to do a pullrequest from the comfort of my command line, without going off to github and forking and all that?
16:04 timotimo TimToady: yup, get the "hub" program
16:05 timotimo masak: kindly look at your pull requests for Adventure::Engine and Game::Crypt :)
16:05 timotimo and yesterday i made Druid work again, too
16:05 timotimo since we're about to have a Rakudo Star release, it's a good time to be fixing ecosystem modules
16:10 timotimo i can't find the huge log samcv made where it shows what modules failed to install?
16:10 timotimo oh, is this it
16:10 timotimo http://hack.p6c.org/~samcv/install_log.txt
16:11 timotimo masak: you think it might be time to throw the GGE out of the ecosystem?
16:11 robertle joined #perl6
16:13 mcmillhj joined #perl6
16:18 Cabanossi joined #perl6
16:26 astj joined #perl6
16:30 El_Che if someone here bumps agains the travis timeout my solution is here: https://github.com/nxadm/rakudo-pkg/blob/master/.travis.yml
16:31 mcmillhj joined #perl6
16:45 mcmillhj joined #perl6
16:45 domidumont joined #perl6
16:49 TreyHarris TimToady: if you run into any issues just holler; I've set up hub and dealt with many issues on many different platforms for my companies' devs over the years
16:51 timotimo m: say (1..10).map({ { :a($_) } }); # what is preventing the inner block from being parsed as a block? it has $_ in it …
16:51 camelia rakudo-moar d27266: OUTPUT: «(a => 1 a => 2 a => 3 a => 4 a => 5 a => 6 a => 7 a => 8 a => 9 a => 10)␤»
16:51 TreyHarris Erm, can't run mi6 anymore: https://gist.github.com/treyharris/bc343feb1162be5cf79e4c98acc2da40
16:52 timotimo TreyHarris: huh, what happened to that file? it's got some powershell stuff in it?
16:52 TreyHarris I'm on ubuntu
16:52 timotimo yeah but %~dpn0 is powershell
16:53 TreyHarris I installed it with zef, maybe retry w/panda? People have been telling me to use zef
16:53 timotimo no zef is right
16:54 timotimo that's literally all there is in the file
16:54 timotimo um
16:54 timotimo it's installing a windows launcher
16:55 TreyHarris the newest commits to zef are windows-related
16:55 timotimo 2 bin/ scripts [mi6 mi6.bat] installed to:
16:55 timotimo /home/timo/perl6/install/share/perl6/site/bin
16:56 timotimo but only one of the two actually exist
16:56 timotimo my bet is that it accidentally installed mi6.bat over mi6 for some reason
16:56 TreyHarris nope, no they aren't, I don't know where I saw recent win-related commits
16:56 timotimo for the time being, just clone mi6 locall, and rm bin/mi6.bat, then zef install --force .
16:58 TreyHarris timotimo++ thanks
16:59 japh__ joined #perl6
17:00 timotimo i don't think you should ever have a bat in your bin/
17:00 timotimo rakudo or zef or whoever makes a launcher for you if you're on windows
17:03 mcmillhj joined #perl6
17:04 TreyHarris does the bat file let an author on windows test directly before install time? I know nothing of windows except the see-through kind and the orbital kind
17:05 timotimo nah, you'd just perl6 -Ilib bin/mi6 in that case
17:06 timotimo oh man, having fought cmake today left me in a surpremely foul mood
17:07 TreyHarris timotimo: try dealing with an airline and aviation lawyers, I bet after that cmake would seem like a soothing hottub
17:08 timotimo i know i could always have it worse
17:08 timotimo trying to develop krita is a hundred times more annoying than contributing to pypy. and pypy has a one-hour compile cycle.
17:12 mcmillhj joined #perl6
17:15 bwisti joined #perl6
17:16 sufrostico joined #perl6
17:33 Cabanossi joined #perl6
17:34 skids bisect: role A { }; sub A ($) { 42  };  A(2).say # I thought this used to DWIM.
17:34 bisectable6 skids, Bisecting by output (old=2015.12 new=d272667) because on both starting points the exit code is 1
17:34 bisectable6 skids, bisect log: https://gist.github.com/123c5b03ba5697fccf04ccaa416b1ad4
17:34 bisectable6 skids, (2016-05-12) https://github.com/rakudo/rakudo/commit/33ef5a95459b95bf7256391a053550d4efe221de
17:35 titsuki joined #perl6
17:41 skids commit: role A { }; sub A ($) { 42  };  A(2).say # I thought this used to DWIM.
17:41 committable6 skids, ¦role: «Cannot find this revision (did you mean “all”?)»
17:42 jnthn Probably not since the SomeTypeName(...) syntax was taken for coersions
17:42 jnthn *coercions
17:42 skids Isnn't the way to implement coercions to define a sub though?
17:43 skids Oh, coercive types.
17:43 timotimo nah, subs for coercions is no longer a thing
17:43 skids Wait it was postcirumfix, then subs, now what?
17:53 sufrostico joined #perl6
17:57 timotimo postcircumfix is a sub
17:57 timotimo or .. was it built on top of a postcircumfix method?
17:57 skids I think it was a method.
17:57 timotimo now coercions are just methods named the target type
17:57 timotimo which has different problems than the multi sub had
17:57 timotimo but everything that was attempted was problematic
17:58 skids Yeah, noticing that.  Somewhat LTA when you are trying to emulate an enum.
18:02 gdonald joined #perl6
18:09 TreyHarris Every time I see somebody use something like bisectable6 or committable6 I want to shout "this is sorcery, burn the witch!"
18:12 TreyHarris Though at Google I taught git to nooglers and when I got to the bisect section, I saw lots of mouths hanging open. So I guess it's just one of those things you either are used to or seem like pure magic.
18:13 moritz I'm kinda used to it, and still looks like pure magic to me :-)
18:14 moritz speaking of like magic: I'm trying https://github.com/dxa4481/truffleHog right now, and I'm quite impressed so far
18:15 moritz it found faked key material that we use in unit tests (for code that validates the format of the key material)
18:15 moritz and remarkably few false positives
18:17 AlexDaniel joined #perl6
18:18 tony-o anyone familiar with a p6 module connecting to dynamodb?
18:18 yoleaux 23 Nov 2016 15:21Z <jnthn> tony-o: Rakudo 6bb8823 nails, best I can tell, the odd crashes I've been seeing in things using Proc::Async in chars mode.
18:18 tony-o jnthn++
18:19 tony-o though i think we've chatted since nov :-)
18:19 timotimo perhaps in the other channel
18:22 [Coke] /win 6
18:22 timotimo we should have a bot that collects lines of /win N and draws a winning number at the end of each week
18:26 geekosaur .oO ( and the winner is: bal )
18:26 El_Che new rakudo linux packages: https://github.com/nxadm/rakudo-pkg/releases/tag/2017.04.2
18:29 timotimo cool
18:41 ufobat joined #perl6
18:43 sufrostico joined #perl6
18:43 iyra joined #perl6
18:44 RabidGravy Boom!
18:45 iyra joined #perl6
18:51 lizmat joined #perl6
18:52 iyra joined #perl6
19:02 Cabanossi joined #perl6
19:03 AlexDaniel joined #perl6
19:03 xinming joined #perl6
19:08 zakharyas joined #perl6
19:17 espadrine joined #perl6
19:32 Cabanossi joined #perl6
19:32 timotimo how's it going, gravy sempai?
19:33 samcv good **
19:33 timotimo good HyperWhatever to you, too
19:33 gdonald joined #perl6
19:35 RabidGravy Harr, samcv - I'm just testing all of the ecosystem with your licence test
19:36 timotimo samcv: i sent out three pullrequests to fix three modules in the ecosystem today :)
19:36 samcv oh nice
19:37 RabidGravy also, rather than "meta-version" there is an already specced "meta6"
19:37 samcv RabidGravy, tell me what you find. probably gonna have almost 108 failures due to licenses. since some had "Artistic" when their project had Artistic 2.0, or put Artistic2 or Artistic 2 or they have the url to the license
19:37 samcv yeah i saw that RabidGravy
19:37 samcv thank-
19:41 samcv interestigly thre are 3 different artistic 1.0 licenses on the spdx website |X confusing. at least there's only one Artistic-2.0
19:42 timotimo clearly Artistic-2.0 was made because there were too many different Artistic-1.0s
19:42 samcv was that why? i thought it was other reasons
19:43 timotimo i have no idea :)
19:43 samcv :)
19:43 timotimo i thought it was a funny reference
19:43 MasterDuke_ which xkcd is that? the standards one
19:44 timotimo that one, yes
19:44 timotimo wait, there are too many xkcds about standards
19:44 timotimo we should make a better xkcd
19:44 samcv not sure why there are multiple artistic 1.0 i mean. it is 1.0 it should be the same X|
19:44 samcv lol!
19:44 samcv https://xkcd.com/927/
19:44 labster joined #perl6
19:45 samcv also seems you can view the spdx spec in at least two formats that are the names of animals https://spdx.org/specifications
19:45 samcv what's up with that
19:47 RabidGravy 106 fail
19:48 samcv k. do you have that list?
19:48 samcv or log or something, or were they all due to license failures and not other things?
19:49 japh____ joined #perl6
19:54 RabidGravy samcv, would have failed the licence test https://gist.github.com/jonathanstowe/e8b53b1b180de1c75ada87ab016b69d9
19:54 rindolf samcv: hi! I've collected more links like that xkcd strip here - http://shlomifishswiki.branchable.com/How_Alternatives_Proliferate/?updated
19:54 perigrin joined #perl6
19:57 perigrin joined #perl6
20:00 samcv making lots of PR for projects that have a LICENSE file in the repo, but no field in the META file. progress
20:00 samcv thx for the list RabidGravy
20:00 geekosaur joined #perl6
20:03 samcv time to PR away!
20:04 RabidGravy do not send any for my modules
20:04 RabidGravy and it is spelled LICENCE in all of mine
20:06 samcv in the meta? or the filenames
20:07 timotimo LICHENS
20:07 samcv hehehe
20:07 RabidGravy the filename, being the correct spelling and all
20:07 samcv oh that's fine. but the meta field should be license since that's what the spec says
20:09 RabidGravy using a verb there is obtuse
20:10 samcv in the meta or in the filename or both
20:10 samcv or *anywhere*  heh
20:11 samcv google says licence is a verb and license is a noun
20:11 samcv https://www.google.com/search?q=licence&amp;oq=licence&amp;aqs=chrome.0.69i59j69i60j69i61j69i60.1076j0j4&amp;sourceid=chrome&amp;ie=UTF-8
20:11 samcv with their dictionary thing
20:12 samcv idk though.
20:12 timotimo samcv:  license AE / licence BE
20:12 samcv i don't know what that means
20:12 samcv this ones' license is "The Artistic License 2.0"
20:12 samcv so many variations
20:13 MasterDuke_ it's another british/american thing
20:13 samcv the artistic, artistic, Artistic 2, Artistic-2 so many
20:13 RabidGravy license is a verb, licence a noun yes
20:13 samcv yeah but the json should probably be the standardized thing yes?
20:13 samcv since it's processed by computers
20:13 samcv i don't care what your license file is named, it could be licence.doc but
20:14 samcv as long as the meta file is fine
20:14 RabidGravy and atrribute which isn't boolean should be a vnoun
20:14 RabidGravy maybe I should delete all my modules
20:14 samcv :\
20:15 samcv please don't
20:16 mst given "atrribute" and "vnoun" I'm pretty sure he'll end up typing 'mr' instead of 'rm' anyway
20:16 yoleaux 12:26Z <Zoffix> mst: would you know whether it's possible to debug Travis's IRC bot connection timeouts to Freenode? We've not seen their bot for a while and they responded to me: https://twitter.com/travisci/status/855034311103455232
20:16 mst so we're probably safe
20:17 RabidGravy pfft hippy
20:17 samcv vnoun sounds like a futuristic nounn
20:17 samcv i'm going to start with all packages that depend on Test::META though RabidGravy
20:18 RabidGravy as long as you don't send me any, feel free :)
20:20 samcv :(
20:20 mst if you're going to send RabidGravy a package it generally needs to contain alcohol
20:20 RabidGravy I'm going to make it not fail unless meta6 is 1 or  greater anyway
20:21 samcv you don't want any license field in your meta file RabidGravy even if you have a licence for the project?
20:21 samcv like this https://github.com/jonathanstowe/JSON-Infer/blob/master/META6.json you have a license https://github.com/jonathanstowe/JSON-Infer/blob/master/LICENCE
20:23 RabidGravy it's an optional attribute in the spec
20:24 samcv so you don't want to do any option attributes?
20:24 jnthn Do we have an immediate use for the data?
20:24 samcv well it would be nice if 100% of packages in the ecosystem had license fields
20:25 ranguard setting standards for people to follow?
20:25 samcv and eventually it should probably be required to hav a license to get in the ecosystem
20:25 samcv if you have a non ecosystem module fine have no license field... but if it's in the ecosystem it should have one
20:26 RabidGravy then the spec shouldn't have it as optional
20:26 samcv the meta spec? or ecosystem requirements?
20:26 samcv or either?
20:26 RabidGravy either
20:26 samcv i mean ecosystem is different than the meta spec itself
20:26 * mst is still boggling at travis
20:26 jnthn I'd think the first place to update would be the documentation that we link to from modules.perl6.org: https://docs.perl6.org/language/modules.html#Distributing_Modules
20:27 jnthn Which at present doesn't include license in the template that, in reality, 90%+ of people will just copy/paste/fill out :)
20:27 samcv yeah i need to update the docs there
20:27 samcv so the example has a license field. people need all the reminders they can get. cause some may copy paste the meta6.json on that page and edit it
20:28 samcv and if it starts with no licese field, they're unlikely to add one
20:28 jnthn It's exactly what I'd do if I was new to a community and looking to submit modules
20:28 samcv yeah
20:28 RabidGravy so you are proposing removing modules without a liicense field in the meta?
20:28 samcv i am not RabidGravy
20:29 samcv i am proposing not accepting new additions to the ecosystem without a license field. at some future point in time
20:29 samcv starting out with a gentle nudge to do so when submitting a PR to the ecosystem: https://github.com/perl6/ecosystem/blob/master/PULL_REQUEST_TEMPLATE.md
20:29 samcv but a module without a license, you won't know if you can use it before installing it
20:30 samcv which is non-ideal
20:32 * samcv goes to edit that docs page
20:33 samcv omg. navbar plus got a filtering capability so i can filter routines and methods in the bar! woo
20:33 samcv that's pretty awesome
20:43 samcv also if you have two licenses which at least one thing i saw uses Artistic-1.0-Perl and GPL, it would be `Artistic-1.0-Perl OR GPL-1.0`
20:43 samcv and yes RabidGravy i still need to program in that situation into the meta tester X| don't shoot me!
20:43 TreyHarris samcv: non-ideal, and in several corp environments I've worked at, impossible b/c you need pre-approval for even downloading.
20:43 samcv yes.
20:43 samcv which is why the meta files themselves are under permissible licensing, by submitting to ecosystem, so you can dl the meta files. but dling the module you can't see until you dl...
20:44 samcv and without standardized names, having a whitelist in your package manager (zef supports this) becames useless
20:46 [Coke] I find the licensing of the meta files outrageous, I hope it's actually required.
20:47 [Coke] but absolutely appreciate the effort to get things licensed so I know if I can use them or not.
20:47 * jnthn is, personally, rather doubtful that the META6.json files are copyrightable
20:48 jnthn But yeah, I've been rotten about making sure my modules have a license
20:48 RabidGravy I'm going with removing my modules from the ecosystem until this has some proper discussion
20:49 jnthn So a reminder to do so upon submitting is good, as is the documentation reminding me.
20:49 samcv heh jnthn well i gave you two new issues :)
20:49 samcv i had no liceensed modules either so. i am guilty too
20:49 sjn jnthn: any written work is copyrightable. if you can write it, you can copyright it
20:50 samcv hence my reasoning for https://github.com/perl6/ecosystem/issues/321 [Coke] and TreyHarris
20:51 samcv to have a metadata license that covers the metadata itself. to be more explicit. cause the metadata can be hosted anywhere, it's not like cpan where submitting is controlled, they could replace it at any time with something different that may or may not be copyrighted differently
20:51 samcv (regarding the meta file itself)
20:52 samcv so i take freedesktop.org's recommendations for allowed licenses for metadata and add in artistic 2.0 in addition to that, since i'm mostly certain it fullfillls the requirements
20:52 jnthn sjn: https://cyber.harvard.edu/property00/alternatives/non-copyrightable.html
20:52 gfldex sjn: not all countries got the same laws, here in Germany we got the concept of "Erfindungshöhe" what makes it questionable if porn is copyrightable.
20:53 gfldex what also means we are better safe then sorry by having some sort of ruling on the meta file
20:53 samcv jnthn, united states is not the whole world :P and there are so many different countries
20:53 samcv but that's somewhat encouraging at least. but
20:54 sjn gfldex: isn't that more related to patents? (we have something similar in Norway)
20:54 samcv also what if the person writes like pages in the description of the metadata file?
20:54 samcv like writes a book. and pastes it in the description of the metadata
20:54 gfldex sjn: it applies to any creation (Werk is not easy to translate to english)
20:54 samcv that could be copyrightable easily
20:55 sjn gfldex: the translation is "a Work" (e.g. "a Work of art")
20:55 jnthn samcv: Yes, I know, was just pointing out it's not so clear cut
20:55 gfldex samcv: a poem would do. And given that we are talking about perl folk …
20:55 samcv yep
20:55 samcv hehe
20:56 sjn but sure, I get jnthn's and gfldex' points. there *might* be some expectations of effort before one can claim a ©
20:56 gfldex perl -e 'do not $fear and do not die'
20:57 sjn one way to get around it is by making it easy to auto-generate the META6.json file :)
20:58 gfldex already done
21:00 samcv “ Four of these were fictitious listings that Rural had inserted into its directory to detect copying.”
21:00 samcv hehehe
21:00 samcv putting fake phone numbers and names in. sneaky
21:00 gfldex they do that with maps too
21:01 samcv yea
21:01 samcv does google maps do that though?
21:01 samcv and how many poeple have been in an accident due to inaccurate listings caused by copy protection and has anybody sued heh
21:01 samcv https://en.wikipedia.org/wiki/Trap_street
21:01 gfldex may become dangerouse with self driving cars
21:03 samcv trap streets aren't copyrightable aparently. but i guess they are evidence that the other sections are copied
21:03 samcv There, the court stated: "[t]o treat 'false' facts interspersed among actual facts and represented as actual facts as fiction would mean that no one could ever reproduce or copy actual facts without risk of reproducing a false fact and thereby violating a copyright ... If such were the law, information could never be reproduced or widely disseminated." (Id. at 733)
21:03 samcv i like that decision!
21:04 dct joined #perl6
21:07 TEttinger joined #perl6
21:09 kyan joined #perl6
21:12 itaipu joined #perl6
21:17 bjz joined #perl6
21:20 japh__ joined #perl6
21:22 sufrostico joined #perl6
21:34 lizmat joined #perl6
21:48 Cabanossi joined #perl6
22:06 samcv on reading the do what the fuck you want to license again. it's really bad. it doesn't even specify that the license relates to the softwares use or otherwise
22:06 samcv oh. i guess version 2 does sorta
22:07 samcv DO WHAT THE FUCK YOU WANT TO PUBLIC LICENSE
22:07 samcv TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
22:07 samcv 0. You just DO WHAT THE FUCK YOU WANT TO.
22:07 samcv still not very clear even in version 2.0. sure you can always do WTF you want to. i mean. but legally. it gives you nothing
22:08 samcv i would argue that you *can* always do wtf you want to, and people already do. but it only says what you can do, not that the copyright holder is actually granting you those privledges
22:20 samcv the CC-zero actually dedicates to public domain with plenty of clauses that take effect if you cannot actually reliquish copyright in your or some other country and in that case grants licenses to all persons, if the other sections are not fully valid
22:20 samcv pretty nice
22:21 timotimo yeah the CC people know their stuff
22:22 itaipu joined #perl6
22:23 samcv gfldex, i made a zef issue open for comments on the NONE vs NOASSERTION thing and mentioned what you suggested https://github.com/ugexe/zef/issues/154
22:23 samcv if anybody wants to comment on it
22:28 iyra joined #perl6
22:29 iyra what's the best way to modify a hash/array in perl 6 from another function? I want to add values to a hash for example from the function I am passing it to
22:29 timotimo you can just assign to subscripts, or use the push method or sub
22:29 samcv iyra, hashes are able to be written to by default
22:30 timotimo if you want a read-only hash, you'd use Map
22:30 samcv m: sub thing ( %hash) { %hash<blah> = 'foo' }; my %hash = 'abc' => '123; thing(%hash); say %hash
22:30 camelia rakudo-moar ce37b3: OUTPUT: «5===SORRY!5=== Error while compiling <tmp>␤Unable to parse expression in single quotes; couldn't find final "'" ␤at <tmp>:1␤------> 3= 'abc' => '123; thing(%hash); say %hash7⏏5<EOL>␤    expecting any of:␤        single quotes␤      …»
22:30 samcv m: sub thing ( %hash) { %hash<blah> = 'foo' }; my %hash = 'abc' => '123'; thing(%hash); say %hash
22:30 camelia rakudo-moar ce37b3: OUTPUT: «{abc => 123, blah => foo}␤»
22:30 iyra what about arrays? (the things with the @ sigil I think)
22:30 samcv but also push works too
22:30 timotimo same deal with @ sigiled things
22:30 samcv yeah arrays and hashes i believe are writable, sigil things are not unless you use `is rw` timotimo that's correct yes?
22:31 timotimo you mean $ sigil things
22:31 iyra ah, thanks okay
22:31 timotimo "is rw" is one way, "is raw" is another
22:31 samcv yeah that's what i said. sigil things are different than @ or %
22:31 timotimo and things returned from a sub are also made read-only, unless you make the sub itself "is rw" and use either return-rw or last-statement-of-block return
22:32 timotimo @ is a sigil and % is also a sigil :)
22:33 timotimo samcv: are you going to re-generate install_log.txt regularly?
22:33 samcv uh
22:33 samcv i shuold
22:33 samcv i forget what happened at the end that caused it to stall
22:33 timotimo do we have a big checklist like the io checklist zoffix made for the IO grant? where there were checkboxes for "pr submitted" and such?
22:33 samcv yes
22:34 samcv https://github.com/perl6/ecosystem/issues/318 incomplete list
22:34 timotimo aha!
22:34 samcv feel free to add more or edit it
22:35 timotimo wait wtf
22:35 timotimo where did my clone of druid go? with all the fixes?
22:35 timotimo oh!
22:35 timotimo i had push access to druid and pushed them without forking
22:35 samcv i'll regen install_log_live.txt in a bit probably or something. i'll probably implement randomizing install order of modules too
22:36 samcv i should look back in logs and see why it stalled. i have timeouts but. something weird. brb 15 mins
22:36 timotimo what makes the list incomplete? you had to do it by hand?
22:36 samcv uhm
22:36 samcv it's not updated to everything in install_log_live.txt
22:37 samcv it was based on an older log that didn't have the ability to end zef when it timed out
22:37 timotimo OK
22:37 samcv so it was partial
22:37 timotimo that explains it
22:37 samcv also. note that there is an issue. where it'll say OK before instaling certain modules
22:37 samcv so look for the FAIL's not OK cause there's an issue with my script sometimes saying OK before it installs it
22:38 samcv brb
22:39 timotimo OK, looking for FAIL only is easy
22:55 timotimo https://github.com/MoarVM/MoarVM/releases/tag/9999-please-do-not-use-these - perfect
22:55 sammers samcv, looks like Template::Mustache was reported here https://github.com/softmoth/p6-Template-Mustache/issues/16
22:58 astj joined #perl6
22:59 BenGoldberg joined #perl6
23:01 mcmillhj joined #perl6
23:08 kaare_ joined #perl6
23:10 samcv i wonder how many PR i can get done today https://gist.github.com/samcv/b7809f40477812328914fff08efeb6bd
23:11 samcv at least trying to do the more popular ones
23:11 timotimo whoa, you wanna *fix* all of these?
23:11 samcv uh
23:11 timotimo oh
23:11 samcv idk :(
23:11 timotimo no, that's for licenses, right?
23:11 samcv they have licenses, but it's not standard identifier
23:11 samcv the list of no license is bigger than that
23:11 timotimo for a moment i thought this was about the fail-to-install log
23:11 samcv idk can maybe automate PR?
23:12 samcv haha
23:12 samcv dunno
23:12 timotimo you can get the "hub" program and shell-script it
23:12 samcv i've been copy pasting the same commit message and such
23:12 samcv i have hub
23:12 samcv i'm worried about it making a mistake though :X
23:12 samcv plus i have to review all the different ones
23:12 timotimo hmm
23:12 timotimo fair enough
23:12 samcv and connect module names with repos etc. idk.
23:13 timotimo well, for module names and repos you can definitely use modules.perl6.org
23:13 samcv but we can automate opening issues with everybody that has NO license in meta
23:13 samcv yeah
23:13 timotimo i.e. modules.perl6.org/repo/Foo::Bar
23:13 samcv ah
23:13 timotimo will http redirect you
23:13 samcv yeah
23:13 timotimo no* parsing required
23:15 zapwai joined #perl6
23:17 samcv and many people have "Artistic" as their license in the meta, but the project is obviously Artistic 2.0, well the license file in the project says so
23:17 samcv so another reason i don't want to automate. but could automate all the ones that use the artistic license url
23:20 samcv but i have a text file that i have variations of issues or of commit title/descriptions. so
23:20 samcv will see about automating
23:20 timotimo right
23:20 timotimo i'll go to bed now :)
23:20 iyra joined #perl6
23:21 samcv night!
23:21 mcmillhj joined #perl6
23:22 timotimo night!
23:22 skids joined #perl6
23:23 samcv gonna start on some automated thing now. first gonna compile all the different fields of those modules then decide which i can automate and which i cannot
23:27 AlexDaniel TreyHarris: well, 「git bisect」 is just a binary search
23:28 AlexDaniel TreyHarris: so I don't know… what's so magic about binary search? :)
23:31 TimToady samcv: you can probably derive an "auth" from "source-url"
23:32 TimToady looks like a lot of people are using the "github:myname" form of auth
23:32 AlexDaniel TreyHarris: bisectable6 does a bit more sorcery though. Normally 「git bisect」 will take your word on whether the starting points are good or bad, but bisectable6 actually *runs* the code on starting points to figure out what strategy should be used (exit code / signal / output)
23:33 Cabanossi joined #perl6
23:35 TimToady samcv: some people are using "author" to put a name in, but that's kind of different than "auth", so we might leave "author" as is (or change to "creator"?) and go for something unique for auth, github:foo or cpan:bar or an email or something that is unlikely to have collisions, compared to names
23:35 samcv yeah. i think that makes sense
23:36 samcv idk maybe it should create the auth i mean the package manager. cause package manager knows where it got them... otherwise auth is meaningless
23:36 iyra what's a nice way in perl to check if all members of a list return true on a certain predicate? for example, to know if all IO::Paths in an array are directories?
23:36 samcv unless you can verify identity
23:36 TimToady it would also be nice if there were some way to force a new version when a module is patched, but I dunno how to do that...
23:36 samcv if it didn't really come from that github user... then could be malicious
23:36 samcv TimToady, uhm
23:36 AlexDaniel iyra: what about @smth.all ~~ :d
23:36 samcv it's possible. we collect all the META into one json file
23:37 iyra AlexDaniel, thanks, i'll check it out
23:37 samcv apparently. that the zef uses
23:37 samcv but. that. would be hard to do? without scraping
23:37 samcv argh
23:37 AlexDaniel iyra: see https://docs.perl6.org/type/Junction
23:37 samcv TimToady, make so to reinstall ALL of your modules every time!
23:38 TimToady like I sent a pr to jnthn to fix a bug, but forgot to increment the patchlevel, so zef doesn't see it
23:38 TimToady (apparently)
23:39 TimToady the S11 specs require that any given name/ver/auth correspond to a unique text if installed into an official library
23:40 TimToady or we should include a hash as part of the identity, if that's not practical
23:40 samcv hmm
23:40 samcv TimToady, OR we can have uh git revision
23:41 TimToady but it's gonna be a nightmare if you can't name the "correct" version you're looking for uniquely
23:41 samcv in the metadata, like we add it when we collate all of the metadata
23:41 samcv and have it know it has changed
23:41 samcv though the version number will be the same. but realistically. that version number is technically the newest git commit that has that version number in the json...
23:41 samcv so it changes... but. yeah
23:42 TimToady would be nice to have a patchlevel sequence one way or another, whether you believe in semver or not
23:42 samcv since we clone the git when installing a module we could add an extra field or data that denotes the commit hash that the json came from
23:42 samcv yeah
23:45 TimToady but we really need to enforce the invariant somehow that, if you know the identity, you know the exact contents, or we won't be able to have a distribution system that doesn't care where things come from
23:46 TimToady when it comes to source control, we're still kind of in the world of Usenet, where an email address had to know the route it took to get to you
23:46 TimToady we haven't quite reached the goal of Internet email where you have a unique identity that is independent of route
23:48 TimToady that's why "auth" is "auth" and not "author", because it's really supposed to be a naming authority
23:48 samcv yeah
23:56 ugexe the auth has no relation to where the package manager gets the package from
23:57 ugexe and verifying identity is something a recommendation manager / content storage (CPAN, MetaCPAN) could enforce if they wished
23:57 ugexe they could simply refuse to index anything with an auth they couldnt verify
23:58 TimToady right, from the standpoint of Perl 6, there is no requirement that an "auth" of "github:TimToady" be fetched from github; we're only using it to establish a unique id
23:59 TimToady some well-known place where Someone Else is managing the name collisions

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo